Interview
How can AI help us adapt to an evolving climate?
Neara’s managing director, Taco Engelaar, discusses the role AI and digital twins play in aiding adaptation to an evolving climate with Kris Cooper.
Powered by
A network digital twin model. Credit: Neara
Utility providers are among the industries that must explore how to adapt to volatile climate conditions and whether AI can help with that process.
Last year was the hottest on record, and there has also been an increase in the frequency of storms, forest fires and high wind speeds. Such phenomena is causing power outages across the world, but AI is one emerging technology that may be able to help mitigate the impacts of such climate-driven challenges.
Electric utility software company Neara provides an infrastructure modelling platform that creates digital twins of networks to help utility providers anticipate threats to their services.
Speaking to Energy Monitor, the company’s managing director and senior vice-president, Taco Engelaar, explains how technologies such as those it offers could be vital in maintaining infrastructure networks as extreme weather conditions become more frequent.
Kris Cooper: Why is climate change a growing issue for utility networks? What danger does it pose?
Taco Engelaar: As net-zero deadlines loom ever closer, pressure is intensifying on utilities to boost renewables capacity and accelerate the clean energy transition. At COP28, over 100 countries pledged to triple renewables capacity by 2030 – but while recent reports from the International Energy Agency suggest renewables capacity is at an all-time high, there is still a long way to go.
Progress in connecting more renewable energy to the grid is being roadblocked by significant constraints due to ageing infrastructure and slow, heavily siloed approval processes. As a result, much of our renewable energy is lying dormant. There are around 600 renewables projects waiting to join the grid in England and Wales alone.
Meanwhile, extreme weather is increasingly becoming the new norm as climate change gathers pace. Over the past few months, Europe has seen a relentless roster of devastating storms – which have all posed a significant risk to our energy supply. With ageing infrastructure unprepared to withstand the onslaught, the high winds, flooding, wildfires and torrential rain being unleashed during these storms are causing crippling power outages for millions of households.
Utilities face a growing challenge to rapidly adapt and prepare for increasingly violent weather to keep their communities safe and power connected.
Kris Cooper: How can digital twins of utility networks and AI help in maintaining utility networks?
Taco Engelaar: Digital twin models built using AI and digital modelling technology can give utilities the tools to safely monitor and assess their network more efficiently at scale – enabling them to more easily identify areas at risk or in need of repair. They can provide a hyper-accurate virtual representation of an entire network. This can be used by utilities to identify and monitor assets to a much higher degree of accuracy, without needing to enter the field.
All network assets, from pylons right down to individual power lines, can be assessed in granular detail, alongside any surrounding vegetation, structures or other important context. This helps to eliminate blind spots and give utilities insights that go well beyond the scope of what the human eye can take away from any given field visit. As a result, they are able to more quickly and accurately prioritise areas where intervention may be needed and to flag potential clearance risks – such as vegetation that will touch power lines if the wind picks up – before they can cause an issue.
Digital twin models can also be used to simulate the impact of extreme weather events like wildfires or floods before they happen. This enables utilities to predict the risks facing their network in a safe virtual environment and take preventative action to reduce the risk of outages when a storm hits.
Kris Cooper: How are digital twins of utility networks created? What challenges have there been in developing these?
Taco Engelaar: At Neara, our digital twin models are created using AI and digital modelling technology. They combine and analyse multiple high-fidelity data sources – including LiDAR [light detection and ranging], satellite and geospatial data – to map the precise location of a network’s assets and surrounding environment.
A common challenge is that many utilities come in with poor-quality or altogether missing data. Despite utilities’ best efforts, their GIS [geographic information system] data is notoriously error-prone and easily falls out of date; it’s not uncommon for utilities to learn from the first iteration of our model build that 15–20% of their assets are missing or improperly documented. This is where those other data sources come in, whereby in combination they deliver a hyper-accurate representation of the real-life assets.
GIS conflation and correction is often one of the earliest outcomes we see in new deployments. The model itself plays a pivotal role in utilities’ digital transformation journeys as they identify areas where they need to improve data quality depending on the challenges they’re focused on solving. Once they have a clear picture of their assets’ location and configuration, they can make faster, more informed decisions in diagnosing asset health and conducting the wide range of simulation analyses our platform facilitates, depending on their most pressing objectives.
Kris Cooper: How are areas in need of repair identified via simulating the future impact of storms?
Taco Engelaar: When we use a digital twin model to simulate the impact of a storm, it enables us to predict where and how extreme weather is likely to impact the network. For example, before or during a flood, we can use information on the levels of floodwater expected as well as the elevation and soil density of the ground at different points across the network to identify where the water level is likely to come into contact with power lines. This then enables the utility to prevent damage by switching off these sections of the network in advance and to more quickly and accurately assess when it is safe to restore power after the event.
For example, during a once-in-50-years flooding event, Australian utility company Endeavour Energy used the Neara platform to virtually model the storm. This enabled them to predict areas in particular danger – for example, where there was the biggest risk of clearance limits being breached – and turn off parts of the network at the highest risk from the flood-water. After the flood, they were able to use the model to speed up the restoration of power for people most in need. Use of the digital twin simulation helped to eliminate around 300 hours of manual inspection time.
Kris Cooper: What utility companies is Neara working with?
Taco Engelaar: Neara’s platform is currently being used by utility operators to model more than one million square miles and eight million assets globally. Originally founded in Australia, we’ve since expanded to both the US and Europe and are already partnering with a number of European utilities including ESB in Ireland, HEDNO in Greece and Scottish Power in the UK.
Kris Cooper: How else can AI help adaptation to the changing climate?
Taco Engelaar: As well as helping us to protect our energy infrastructure against the changing climate, AI is being developed to help us weather the impact of climate change across many other areas of our lives.
A few examples include improving forecasting to give more advanced warnings before extreme weather hits, building ‘smart sewers’ to reduce the risk of pollution and flooding, and accelerating the development of climate-resilient crops to maintain food security in the face of more frequent heatwaves and droughts.
